Sexual dimorphism: increased sterol excretion leads to hypocholesterolaemia in female hyperbilirubinaemic Gunn rats

Abstract

Circulating bilirubin is associated with reduced serum cholesterol concentrations in humans and in hyperbilirubinaemic Gunn rats. However, mechanisms contributing to hypocholesterolaemia remain unknown. Therefore, this study aimed to investigate cholesterol synthesis, transport and excretion in mutant Gunn rats. Adult Gunn and control rats were assessed for daily faecal sterol excretion using metabolic cages, and water was supplemented with [1-¹³ C]-acetate to determine cholesterol synthesis. Bile was collected to measure biliary lipid secretion. Serum and liver were collected for biochemical analysis and for gene/protein expression using RT-qPCR and western blot, respectively. Additionally, serum was collected and analysed from juvenile rats. A significant interaction of sex, age and phenotype on circulating lipids was found with adult female Gunn rats reporting significantly lower cholesterol and phospholipids. Female Gunn rats also demonstrated elevated cholesterol synthesis, greater biliary lipid secretion and increased total faecal cholesterol and bile acid excretion. Furthermore, they possessed increased hepatic low-density lipoprotein (LDL) receptor and SREBP2 expression. In contrast, there were no changes to sterol metabolism in adult male Gunn rats. This is the first study to demonstrate elevated faecal sterol excretion in female hyperbilirubinaemic Gunn rats. Increased sterol excretion creates a negative intestinal sterol balance that is compensated for by increased cholesterol synthesis and LDL receptor expression. Therefore, reduced circulating cholesterol is potentially caused by increased hepatic uptake via the LDL receptor. Future studies are required to further evaluate the sexual dimorphism of this response and whether similar findings occur in females with benign unconjugated hyperbilirubinaemia (Gilbert's syndrome). KEY POINTS: Female adult hyperbilirubinaemic (Gunn) rats demonstrated lower circulating cholesterol, corroborating human studies that report a negative association between bilirubin and cholesterol concentrations. Furthermore, female Gunn rats had elevated sterol excretion creating a negative intestinal sterol balance that was compensated for by elevated cholesterol synthesis and increased hepatic low-density lipoprotein (LDL) receptor expression. Therefore, elevated LDL receptor expression potentially leads to reduced circulating cholesterol levels in female Gunn rats providing an explanation for the hypocholesterolaemia observed in humans with elevated bilirubin levels. This study also reports a novel interaction of sex with the hyperbilirubinaemic phenotype on sterol metabolism because changes were only reported in females and not in male Gunn rats. Future studies are required to further evaluate the sexual dimorphism of this response and whether similar findings occur in females with benign unconjugated hyperbilirubinaemia (Gilbert's syndrome).

Keywords: Gilbert's syndrome; LDL receptor; SREBP; UGT1A1; bile acid metabolism; cholesterol metabolism; lipid; unconjugated bilirubin.

Figure 1. De novo fractional cholesterol synthesis and hepatic cholesterol content of adult Gunn (hyperbilirubinaemic) and control (normobilirubinaemic) rats

A, the rate of serum fractional cholesterol synthesis measured as a percentage (%) of newly formed ¹³C‐cholesterol at steady state. B, total cholesterol content per gram of liver tissue. Data are presented as means (SD). Two‐way ANOVA was performed with main effects: phenotype (Gunn or control) and sex (male or female). All post hoc analyses compared differences between phenotypes within the same sex. Statistically significant (P < 0.05) P values are highlighted in bold.

Figure 2. Biliary lipid secretion of adult Gunn (hyperbilirubinaemic) and control (normobilirubinaemic) rats

A, bile flow rate normalized for body weight. B–D, total cholesterol, phospholipids and bile acids secreted through bile normalized for body weight. E, biliary lipid (cholesterol + phospholipid) relative to bile acid secretion (mol:mol). Data are presented as means (SD). Two‐way ANOVA was performed with main effects: phenotype (Gunn or control) and sex (male or female). All post hoc analyses compared differences between phenotypes within the same sex. Statistically significant (P < 0.05) P values are highlighted in bold.

Figure 3. Daily faecal sterol excreted, normalized for body weight of adult Gunn (hyperbilirubinaemic) and control (normobilirubinaemic) rats

A and B, daily faecal excretion of cholesterol normalized for body weight. C and D, daily faecal excretion of bile acids normalized for body weight. E and F, daily faecal excretion of total sterols (cholesterol + bile acids) normalized for body weight. Data are presented as means (SD). Two‐way ANOVA was performed with main effects: phenotype (Gunn or control) and sex (male or female). All post hoc analyses compared differences between phenotypes within the same sex. Statistically significant (P < 0.05) P values are highlighted in bold.

Figure 4. Net intestinal cholesterol flux of adult Gunn (hyperbilirubinaemic) and control (normobilirubinaemic) rats

A, model describing the four mechanisms (diet, biliary secretion, cholesterol reabsorption and TICE) that affect the rate of faecal cholesterol excretion. B, the daily rates of biliary cholesterol secretion, net intestinal cholesterol flux and faecal cholesterol excretion. Since the diet did not contain cholesterol, its contribution was disregarded. The net intestinal cholesterol flux was defined as the overall contribution of TICE and cholesterol reabsorption, and it was estimated by subtracting daily biliary cholesterol secretion from daily faecal cholesterol excretion. C, the daily net intestinal cholesterol flux in a graphical format. Data are presented as means (SD). Two‐way ANOVA was performed with main effects: phenotype (Gunn or control) and sex (male or female). All post hoc analyses compared differences between phenotypes within the same sex. Statistically significant (P < 0.05) P values are highlighted in bold.

Figure 5. Relative composition of biliary bile acid species of adult Gunn (hyperbilirubinaemic) and control (normobilirubinaemic) rats

Bile acid species are presented as a mole percentage (%) of total bile acids excreted over an hour. ‘Others’ represents the sum contribution of A‐MCA, GUDCA, CDCA, GDCA, TLCA, GLCA, β‐MCA, O‐MCA and HDCA. Two‐way ANOVA was performed with main effects: phenotype (Gunn or control) and sex (male or female). All post hoc analyses compared differences between phenotypes within the same sex and only significant (P < 0.05) P values are reported on the figure. [Colour figure can be viewed at wileyonlinelibrary.com]

Figure 6. Relative contribution of biliary glycine and taurine bile acid species and the hydrophobicity index of adult Gunn (hyperbilirubinaemic) and control (normobilirubinaemic) rats

A, bile acid species are presented as a mole percentage (%) of total bile acids excreted over an hour. B, Heuman index of hydrophobicity of the biliary bile acid pool. Two‐way ANOVA was performed with main effects: phenotype (Gunn or control) and sex (male or female). All post hoc analyses compared differences between phenotypes within the same sex. Statistically significant (P < 0.05) P values are highlighted in bold.

Figure 7. Hepatic gene expression of female adult Gunn (hyperbilirubinaemic) and control (normobilirubinaemic) rats

A, genes that were significantly different between groups (P < 0.05). B, genes that tended (P < 0.10) to demonstrate differences. Abbreviations: Pmvk, phosphomevalonate kinase; Cyb5r3, NADH‐cytochrome b5 reductase 3; Lrp6, low‐density lipoprotein receptor‐related protein 6; Colec12, collectin12; Apof, apolipoprotein F; Soat2, sterol O‐acyltransferase 2; Ldlr, low‐density lipoprotein receptor; Scap, Srebf cleavage activating protein; Tm7sf2, transmembrane 7 subfamily member 2; Trerf1, transcriptional regulating factor 1; Nr1h2, nuclear receptor subfamily 1, group H, member 2; Insig1‐2, insulin‐induced gene 1–2; Srebf2, sterol regulatory element‐binding factor 2. A two‐tailed unpaired t test was used for statistical analysis. Fold change data from Table 4 were transformed to log₂ for graphical purposes. Data are presented as medians ± range.

Figure 8. Protein expression analysis using western blot for targets of cholesterol synthesis, transport and breakdown in livers from adult Gunn (hyperbilirubinaemic) and control (normobilirubinaemic) rats

A–E, HMGCR, CYB5R3, CYP7A1 and ABCA1 expression normalized to GAPDH expression and LDLr expression normalized to B‐actin expression. F, cropped image that represents a single western blot run. Abbreviations: HMGCR, 3‐hydroxy‐3‐methyl‐glutaryl‐coenzyme A reductase; CYB5R3, NADH‐cytochrome b5 reductase 3; CYP7A1, cholesterol 7 alpha‐hydroxylase; LDLr, low‐density lipoprotein receptor; ABCA1, ATP‐binding cassette transporter; Data are presented as means (SD). A two‐tailed unpaired t test was used for statistical analysis. Statistically significant (P < 0.05) P values are highlighted in bold.

Figure 9. Protein expression analysis using western blot for nuclear form of SREBP2 (n‐SREBP2) in livers from adult female Gunn (hyperbilirubinaemic) and control (normobilirubinaemic) rats

A, SREBP2 expression normalized to GAPDH expression. B, an example cropped image that represents a single western blot run. Abbreviations: SREBP2, sterol regulatory element‐binding protein 2. Data are presented as means (SD). Statistical analyses were conducted using unpaired t tests. Statistically significant (P < 0.05) P values are highlighted in bold.